Where does AI get its facts?

Check out this chart 👇 — AI models like ChatGPT and Perplexity are pulling info from all over the internet.

Reddit (40.1%) → crowd wisdom, hot takes, and the occasional chaos.
Wikipedia (26.3%) → the internet’s encyclopedia… but editable by anyone.
YouTube & Google (23%+) → video breakdowns and search results.
Yelp (21%) → yes, even that “one-star” review about cold fries might feed an AI.

Here’s the beauty: AI pulls from the hive mind, not just ivory-tower textbooks. That’s why it feels conversational, human, and sometimes hilariously off-base.

⚡ Greatness: AI synthesizes perspectives from millions of voices. It can be more current, more nuanced, and sometimes more real than a single expert.

⚠️ Pitfalls: If the source is flawed, biased, or just plain wrong… AI can echo that too. Garbage in = garbage out.

👉 Question for you all:Do you trust AI more when it draws from the “crowd” (Reddit, Yelp) or from “structured” sources (Wikipedia, Google, maps)?

Drop your take below — curious to see where this community lands. I have specified to CHAT GPT where I want it to pull data from and have yielded different answers.

#AItrust

9 comments